The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes output from git log and git diff to automatically draft pull request titles and descriptions. This introduces a risk where malicious instructions hidden in code comments or commit messages could influence the agent's behavior during the drafting process.
Ingestion points: In SKILL.md, the agent is instructed to read output from git log <base>..HEAD and git diff <base>...HEAD.
Boundary markers: There are no specific delimiters or negative constraints provided to the agent to ignore instructions that might be contained within the diff or log content.
Capability inventory: The skill has the ability to execute git push and gh pr create to modify remote repository state.
Sanitization: The skill does not include steps to sanitize or filter the content of commit messages or code changes before they are summarized.

pull-request